ATBECHDEDKESFRGBGRITLILUNLSEMCPTIESILTLVFIROMKCYALTRBGCZEEHUPLSK..HRIS..MTNORS..SM..................JBDM Ver 0.1.67 (18 Oct 2017) - 2100000/03448052EUROPEAN PATENT SPECIFICATIONB120190821EP17849890.32017071320180323zhenen20171464872720170619CN20190821201934201902272019092019082120193420190401H04N 21/845 20110101AFI20180627BHEP H04N 21/24 20110101ALI20180627BHEP H04N 21/442 20110101ALI20180627BHEP H04N 21/8547 20110101ALI20180627BHEP H04N 21/2387 20110101ALI20180627BHEP deVERFAHREN ZUR VERARBEITUNG VON STREAMING- MEDIENDATEIEN UND LIVEÜBERTRAGUNGSSYSTEMenMETHOD FOR PROCESSING STREAMING MEDIA FILE AND LIVE BROADCAST SYSTEMfrPROCÉDÉ DE TRAITEMENT DE FICHIER DE CONTENU MULTIMÉDIA, ET SYSTÈME DE DIFFUSION EN DIRECTCN-A- 101 350 740CN-A- 101 471 919CN-A- 102 098 571CN-A- 106 412 719US-A1- 2015 229 688US-A1- 2017 171 276US-B2- 8 555 13020180703CHEN, HongBuilding A Floor 5 Guangqi Culture Plaza No.2899 Xietu Road Xuhui DistrictShanghai 200030CNLV, ShibiaoBuilding A Floor 5 Guangqi Culture Plaza No.2899 Xietu Road Xuhui DistrictShanghai 200030CNWangsu Science & Technology Co., Ltd.101678766H56-1869-16EPBuilding A Floor 5 Guangqi Culture Plaza 2899 Xietu Road Xuhui DistrictShanghai 200030CNHanna Moore + Curley101668780Garryard House 25/26 Earlsfort TerraceDublin 2, D02 PX51IEALATBEBGCHCYCZDEDKEEESFIFRGBGRHRHUIEISITLILTLULVMCMKMTNLNOPLPTRORSSESISKSMTRCN201709277320170713zhWO201823279620181227201852 FIELD OF THE DISCLOSURE

The present disclosure generally relates to the live streaming technology and, more particularly, relates to a streaming media file processing method and a live streaming system.

BACKGROUND

As Internet is widely used, the need of transmitting audio-video signals over Internet is increasing. Radio, television, and other media are moved to online to distribute the audio-video programs over the Internet. However, when the audio-video programs are stored, the sizes of audio-video files are often huge. Under the circumstance of still limited network bandwidth, spending tens of minutes or even longer waiting for transferring an audio or video file is not a desired user experience.

The introduction of streaming media technology eases the burden of transferring audio-video contents over Internet to a certain extent.

Using the streaming media technology to transfer audio, video, or animation contents from a server to a user computer may be continuous and uninterrupted. The user may not need to wait for the completion of downloading an entire file. After a few seconds or over ten seconds of delay at the beginning, the user may view the content. When the audio-video content is played on the user's device, the remaining portion of the file may continue to be downloaded from the server.

Currently, the streaming media technology is widely used in on-demand services. That is, the audio-video file has already been stored on a server. While the content of the file is played, a user may drag to play at any position of the content of the file. To support the drag-to-play function, the streaming media file needs to be pre-processed. For example, the streaming media file may be marked with streaming timestamps. Finding a position in the content file may be conducted through locating a streaming timestamp in the streaming content file. Alternatively, the streaming media file may be sliced by key image frames such that user may find a position more precisely in the streaming content file. The server may rapidly jump to the requested position in response to the user's request.

However, the described pre-processing methods of the streaming media files may not be suitable for live streaming. When viewing live streaming images, user may not be able to perform the drag-to-play operation on the live streaming content. The user's device may receive the live streaming content in real-time from a source terminal or a server, and may self-adjust according to the status of the user's player client.

As we all know, the real-time requirement for playing audio-video content is high in the live streaming. How to process media files to provide desired viewing experience to the client may be a problem that requires a solution urgently. US 2015/229688 A1 discloses obtaining an index file of a media stream. The media stream includes the index file and at least two slice files. US 2017/171276 A1 discloses slicing a streaming media file. The slicing result carries sequence number information of fragmentation and fragmentation duration information.

BRIEF SUMMARY OF THE DISCLOSURE

To solve issues in the existing technology, the present disclosure provides a streaming media file processing method, comprising: slicing a streaming media file into slices according to a preset time unit; creating a playing sequence number for each slice based on time data, where the playing sequence number is configured to determine an order of playing the slices; and storing the slices and the playing sequence numbers of the slices separately and establishing a mapping relationship between the slices and the corresponding playing sequence numbers.

Optionally, slicing the streaming media file into slices according to the preset time unit includes: obtaining a duration for each frame in the streaming media file; and based on the duration and the preset time unit, obtaining the slice by slicing the streaming media file into the slices.

Optionally, the time data includes a streaming timestamp of a first frame in the slice.

Optionally, creating a playing sequence number for the slice based on the time data includes: applying the time data into a predetermined function to calculate the playing sequence number based on the time data.

Optionally, creating a playing sequence number for the slice based on the time data further includes: determining whether a calibration is required for a currently created playing sequence number; when it is determined that a calibration is required, creating the playing sequence number for the current slice according to the system time; and when it is determined that no calibration is required, applying the time data into a predetermined function to calculate the playing sequence number.

Optionally, determining whether calibration is required for currently created playing sequence number includes: determining whether loss of data occurs between a current slice and a preceding slice; and when it is determined that loss of data has occurred, determining that a calibration is required.

Optionally, playing the steaming media file includes: retrieving a latest playing sequence number; determining whether pausing occurs based on a playing sequence number of the currently played slice and the latest playing sequence number; when it is determined that the pausing has occurred, determining a next playing sequence number based on a playing delay and the latest playing sequence number; when it is determined that no pausing has occurred, determining the next playing sequence number based on the order of playing sequence numbers; and downloading a slice corresponding to the next playing sequence number.

Optionally, determining whether pausing occurs includes: calculating a numerical difference between the playing sequence number of the currently played slice and the latest playing sequence number; and determining whether pausing occurs by determining whether the numerical difference is within a preset range.

Optionally, after the streaming media file is sliced into slices, recording whether each of the slices includes a key frame.

Another aspect of the present disclosure provides a live streaming system, comprising a push-stream terminal, a server, and a player. The push-stream terminal transfers a streaming media file to the server. A file processing module of the server receives the streaming media file, slices the received streaming media file into slices according to a present time unit, creates a playing sequence number for each divided slice based on time data, establishes mapping relationship between the slices and the corresponding playing sequence numbers, and respectively sends the slices, the playing sequence numbers of the slices, and the mapping relationship to a storage module of the server. The storage module stores the slices in a first section, and the playing sequence numbers of the slices and the mapping relationship in a second section. The player obtains and selects the playing sequence number from the second section, and downloads and plays a slice from the first section based on the playing sequence number and the mapping relationship.

Optionally, slicing the streaming media file into slices by the file processing module of the server according to a present time unit includes: obtaining a duration for each frame in the steaming media file; and obtaining the slice by slicing the streaming media file based on the frame durations and the preset time unit.

Optionally, the time data includes a streaming timestamp of a first frame in the slice.

Optionally, creating a playing sequence number for the slice by the file processing module of the server based on time data includes: applying the time data into a predetermined function to calculate the playing sequence number based on the time data.

Optionally, creating a playing sequence number for the slice by the file processing module of the server based on time data further includes: determining whether a calibration is required for a currently created playing sequence number; when it is determined that a calibration is required, creating the playing sequence number for the current slice according to the system time; and when it is determined that no calibration is required, applying the time data into a predetermined function to calculate the playing sequence number.

Optionally, determining by the file processing module whether a calibration is required for the currently created playing sequence number includes: determining whether loss of data occurs between a current slice and a preceding slice; and when it is determined that loss of data occurred, determining that a calibration is required.

Optionally, the player includes a request unit and a playing unit; after a next playing sequence number is determined, the request unit sends a request to the storage module of the server to download the next slice and sends the downloaded slice to the playing unit to play; and when the request unit determines the playing sequence number for the requested slice, the request unit first obtains the latest playing sequence number, and then determines whether pausing occurs based on the playing sequence number of the currently played slice and the latest playing sequence number; when it is determined that the pausing occurred, the request unit determines a next playing sequence number based on a playing delay and the latest playing sequence number; when it is determined that no pausing occurred, the request unit determines a next playing sequence number based on the order of the playing sequence numbers, and sends a request to the server to download the slice corresponding to the next playing sequence number.

Optionally, the request unit calculates a numerical difference between the playing sequence number of the currently played slice and the latest playing sequence umber, and determines whether pausing occurs based on whether the numerical difference is within a preset range.

Optionally, after the file processing module divides the streaming media file into slices, whether each of the slices includes a key frame is recorded.

Optionally, the live streaming system includes a plurality of servers and a plurality of players.

Optionally, the live streaming system includes a plurality of push-stream terminals; and each server is connected to one or more push-stream terminals, respectively.

The streaming media file processing method according to the present disclosure divides the streaming media file into slices according to a time unit, and creates a playing sequence number for each slice based on time data. In the subsequent playing process, the player operates based directly on the playing sequence numbers, and selects suitable slices to play such that when pausing occurs, the player rapidly resumes playing.

In addition, by introducing a calibration step, when the file data is lost in the transmission, the method timely adjusts the playing sequence number such that the correct playing position is located precisely in the subsequent playing process. Thus, the playing request data errors due to loss of data and the incorrect playing sequence number are prevented.

Further, after the streaming media file is divided into slices, each slice that includes a key frame is recorded. When pausing occurs and slice skipping is required, a slice that includes a key frame is facilitated to be quickly identified and played. Thus, the snowy screen symptom when the slice that does not include a key frame is played is prevented, and the user experience is improved.

In the live streaming service, the real-time requirement is important. However, loss of data due to network fluctuations and playing pausing are the problems that require immediate solutions in the live streaming service. The embodiments provided by the present disclosure effectively detect loss of data, take remedy actions accordingly, and rapidly resume from the playing pausing. Thus, the requirement for live streaming service is better satisfied.

BRIEF DESCRIPTION OF THE DRAWINGS

To more clearly illustrate the technical solution in the present disclosure, the accompanying drawings used in the description of the disclosed aspects are briefly described hereinafter. Obviously, the drawings described below are merely some aspects of the present disclosure.

DETAILED DESCRIPTION

Reference will now be made in detail to exemplary embodiments of the disclosure, which are illustrated in the accompanying drawings. Those skilled in the art may readily understand other advantages and effects of the present disclosure as described in the specification.

It should be noted that the illustrations provided in the disclosed embodiments are intended to provide only the basic concept of the present disclosure. Only the components related to the present disclosure are shown in the schematic drawings rather than the actual number, shape or size of the components in the real life applications.

Aspects of the present disclosure are described as follows with references to the accompanying drawings.

FIG. 1 illustrates a flow chart of an exemplary streaming media file processing method according to the present disclosure. As shown in FIG. 1, the streaming media file processing method according to the present disclosure may include the following steps.

Step S10: slicing a streaming media file into slices according to a preset time unit.

Specifically, the streaming media file may essentially comprise an audio-video file that is distributed through streaming. The audio-video stream may include consecutive frames. The consecutive frames may include key frames and intermediate frames. Each frame may include a streaming timestamp and a duration to indicate a starting time to play in the audio-video stream and a playing length of the frame. For example, a first frame in a video stream may have a streaming timestamp 0, and a duration 40ms; a second frame may have a streaming timestamp 40; a third frame may have a streaming timestamp incremented sequentially.

The preset time unit may be used to determine the length of each slice. When the streaming media file is sliced, a slice including a plurality of consecutive frames may be intercepted according to the duration of each frame and the preset time unit. For example, the first frame may have a streaming timestamp 0 and a duration 40ms. The second frame may have a streaming timestamp 40 and a duration 60ms. The third frame may have a streaming timestamp 100. When the preset time unit is 100ms, the slice may include the first frame and the second frame. Therefore, playing time for a single slice may be a sum of durations of all the frames in the slice, which is theoretically equal to the preset time unit.

The preset time unit may be set according to actual requirements. The preset time unit may be a specific value or a range of values, which is not limited by the present disclosure. It should be understood that, because the duration of each frame is determined according to the specific audio-video stream, in the process of slicing, the boundary may not correspond to a frame perfectly. In this case, a rounding method may be set forth to take care of the scenario. For example, a frame closest to the boundary may be selected as the ending frame of the slice. No limitation is imposed by the present disclosure.

Step S20: creating a playing sequence number for the slice based on time data.

Specifically, the time data may include the streaming timestamp of the first frame in the slice. Creating the playing sequence number may include calculating a playing sequence number based on a function. For example, in one embodiment, the calculation function may be: Fx=tx−tx−1+Fx−1, where x is the sequence number of a slice, F(x) is the playing sequence number for the slice x, t(x) is the streaming timestamp of the first frame in the slice x, and F(0) is the current system time.

The created playing sequence number may be used to determine the order of playing the slice. The slice playing sequence number created by the method may intuitively reflect the playing order and specific streaming timestamp of the slices. In the subsequent playing process, the playing sequence numbers may be directly used to play the slices. Further, the playing sequence numbers may be used to detect abnormality, which will be described in detail later.

In certain embodiments, the calculation function may be perfected to include weights and tolerance offset such that the created playing sequence numbers may be more realistic. For example, the calculation function may be Fx=Atx−Btx−1+Fx−1+C, where A and B are weights, and C is a system tolerance offset. The specific values of A, B, and C may be adjusted according to actual scenario of the system design. No limitation is imposed by the present disclosure.

The present disclosure may use other methods to create the playing sequence numbers. For example, the streaming timestamps may be directly used to create the playing sequence numbers. Other time information may be used as long as the created playing sequence numbers reflect the real-time relationship between the slices.

Because the transmission of the streaming media file may be affected by network fluctuation and loss of data may occur, when the occurrence of loss of data is not detected timely, the playing sequence number creation may be affected such that, in the subsequent playing process, the playing sequence numbers may not be obtained correctly.

In one aspect when creating a playing sequence number, the method may first determine whether a calibration is required. When it is determined that a calibration is required, the playing sequence number of the current slice may be created according to the current system time. When it is determined that a calibration is not required, the playing sequence number of the current slice may be created as previously described.

Determining whether a calibration is required includes determining whether loss of data occurs between the current slice and the preceding slice. When loss of data occurs, a calibration may be required.

Specifically, the streaming timestamp of frame may be used to determine whether loss of data occurs. Under normal circumstances, the streaming timestamp difference between adjacent frames of adjacent slices may be the duration of the preceding frame. Frame durations may often be within a reasonable range. By determining whether the streaming timestamp difference between the first frame of the current slice and the last frame of the preceding slice is within a reasonable range, whether loss of data occurs between the current slice and the preceding slice may be determined. In certain other embodiments, whether loss of data occurs may be determined by determining whether the streaming timestamp difference between the first frames of adjacent slices is within a reasonable range.

After it is determined that the loss of data has occurred, when the playing sequence number is still calculated based on the original method, a gap may appear between the playing sequence number of the current slice and the playing sequence number of the preceding slice. In the subsequent playing process, a request may fall into the gap, and no resource may be available to respond to the request. Thus, after it is determined that the loss of data has occurred, a calibration may be required for creating the playing sequence number to fill the gap and to avoid the scenario that no resource is available to respond to certain request. Because the creation of the playing sequence number is based on the system time and the slice playing time, the current system time may be directly used to calibrate the current playing sequence number.

Step S30: storing the slices and the playing sequence numbers of the slices separately and establishing a mapping relationship between the slices and the corresponding playing sequence numbers.

FIG. 2 illustrates a schematic diagram of an exemplary streaming media file processing process according to the present disclosure. As shown in FIG. 2, after being processed in the steps S10 and S20, the streaming media file A may be sliced into a plurality of slices a1, a2, ..., an, a(n+1), ..., each slice may have a corresponding playing sequence number, such as F(1), F(2), ..., F(n), F(n+1), ..., and the slices and the corresponding playing sequence numbers are stored in different storing sections D1 and D2, respectively. Mapping relationship R1, R2, ..., Rn, R(n+1) may be established between the slices and the corresponding playing sequence numbers.

When a slice playing is requested, a player may first determine the playing sequence number of the requested resource, identify the corresponding slice through the mapping relationship, and download the slice. Because it takes time to download the slice, separately storing the slice and the corresponding playing sequence number in different sections may allow independent operations on the slice and the corresponding playing sequence number. Thus, when multiple requests occur at the same time, wait time may be reduced.

The streaming media file processing method including the above-described steps S10 through S30 may slice the streaming media file into slices based on the time characteristics, create the corresponding playing sequence numbers, and timely calibrate the playing sequence numbers in response to loss of data, such that abnormality in the playing process may be handled directly based on the playing sequence numbers. The playing method of the streaming media file that has been processed according to the steaming media file processing method is described in the following.

FIG. 3 illustrates a flow chart of an exemplary streaming media file playing method according to the present disclosure. As shown in FIG. 3, the playing method of the streaming media file may include the following steps.

Step S101: retrieving a latest playing sequence number, where the latest playing sequence number is the playing sequence number that is most recently created and stored.

Step S102: determining whether pausing has occurred based on a playing sequence number of the currently played slice and the latest playing sequence number; when it is determined that pausing has occurred, determining a next playing sequence number based on a playing delay and the latest playing sequence number; when it is determined that no pausing has occurred, determining the next playing sequence number based on the order of the playing sequence number, that is, the playing sequence number immediately following the playing sequence number of the currently played slice.

Specifically, based on the transmission characteristics of the streaming media file, the player may download a subsequent slice for playing at the same time as the current slice is played. In an ideal network environment, the most recently created playing sequence number may be the playing sequence number of the slice for playing next. However, in a real-life network environment, the transmission network may have fluctuations such that the player may not be able to obtain resources timely, and image pausing may occur.

To address the issue, the player may first determine whether the current streaming is pausing to decide the playing sequence number of the slice for playing next. Specifically, determining whether pausing occurs based on the playing sequence number of the currently played slice and the latest playing sequence number may include calculating a numerical difference between the playing sequence number of the currently played slice and the latest playing sequence number, and determining whether pausing occurs based on whether the numerical difference is within a preset range.

Because the playing sequence numbers are created based on time data, the playing sequence numbers may be directly used to determine a playing time for a slice. Theoretically, the numerical difference between the playing sequence number of the currently played slice and the latest playing sequence number may be equal to the playing time of the currently played slice, that is, the preset time unit when the streaming media file is sliced. Further, inevitable playing delay and system tolerance may be accommodated. When the numerical difference is within a reasonable preset range (e.g., the playing time of the slice + playing delay + system tolerance), the currently played slice may be successfully played, and no pausing may occur. Otherwise, the occurrence of pausing may be determined. The playing delay may be the delay in time between the image played at the client terminal and the real-time image at the push-stream terminal due to the time required for processing and transmitting data from the push-stream terminal to the client terminal in live streaming in addition to the delay caused by the equipment reason of the player itself.

When it is determined that no pausing has occurred, the player may directly determine that the next playing sequence number is the sequence number immediately following the currently played sequence number in the record.

When it is determined that the pausing has occurred, to ensure the real-time playing of the images received by the user, the player may skip the delayed slices due to pausing to ensure the real-time playing of the current images. Thus, the slice for playing next may be adjusted. The next playing sequence number may be determined based on the playing delay and the latest playing sequence number. Further, a system tolerance may be accommodated, that is, the next playing sequence number is equal to the latest playing sequence number minus a sum of the playing delay and the system tolerance.

In one aspect in the step S20, after the streaming media file is sliced into slices, each slice that includes a key frame may be recorded. When pausing occurs and image skipping is required, after the next playing sequence number is determined, the player may again determine whether the slice corresponding to the playing sequence number includes a key frame. When it is determined that no key frame is included, a playing sequence number of an adjacent slice that includes a key frame may be selected to be the next playing sequence number. Thus, snowy screen symptom caused by jumping to a slice that does not include a key frame may be avoided.

Step S103: downloading a slice corresponding to the next playing sequence number.

After the next playing sequence number is determined in the step S102, the player may download the slice that corresponds to the next playing sequence number, and may play the downloaded slice after the current slice is completed.

Accordingly, the streaming media file processing method according to the present disclosure may slice the streaming media file into slices based on a time unit, and may create a playing sequence number for each slice based on time data. In the subsequent playing process, the player may operate based directly on the playing sequence numbers, and may select suitable slices to play such that when pausing occurs, the player may rapidly resume playing.

In addition, by introducing a calibration step, when the file data is lost in the transmission, the method may timely adjust the playing sequence number such that the correct playing position may be located precisely in the subsequent playing process. Thus, the playing request data errors due to loss of data and the disorder of the playing sequence numbers may be prevented.

Further, after the streaming media file is sliced into slices, each slice that includes a key frame may be recorded. When pausing occurs and slice skipping is required, the slice that includes a key frame may be facilitated to be quickly identified and played. Thus, the snowy screen symptom when the slice that does not include a key frame is played may be prevented, and the user experience may be improved.

In the live streaming service, the real-time requirement is important. However, loss of data due to network fluctuations and playing pausing are the problems that require immediate solutions in the live streaming service. The embodiments provided by the present disclosure effectively detect loss of data, take remedy actions accordingly, and rapidly resume from the playing pausing. Thus, the requirement for live streaming service is better satisfied.

In addition, the present disclosure also provides a live streaming system.

FIG. 4 illustrates a schematic diagram of an exemplary live streaming system according to the present disclosure. As shown in FIG. 4, the live streaming system 1 may include a push-stream terminal 10, a server 20, and a player 30. The server 20 may be connected to the push-stream terminal 10 and the player 30, respectively.

Specifically, the push-stream terminal 10 may be the source terminal of the audio-video file. In a specific live streaming scenario, the push-stream terminal 10 may be a specific host terminal, or a streaming platform server for a host terminal. In a live streaming, the push-stream terminal 10 may continuously generate the audio-video file, and may transfer the streaming media file of the audio-video stream through streaming to the server 20.

The server 20 may receive the streaming media file from the push-stream terminal 10, and may slice the streaming media file into slices.

FIG. 5 illustrates a schematic diagram of a server of an exemplary live streaming system according to the present disclosure. As shown in FIG. 5, the server 20 may include a file processing module 21 and a storage module 22.

The file processing module 21 may receive the streaming media file, slice the streaming media file into slices according to a preset time unit, create a playing sequence number for each slice based on time data, establish a mapping relationship between each slice and the corresponding playing sequence number, and respectively send the slices, the corresponding playing sequence numbers, and the mapping relationship to the storage module 22.

Specifically, as previously described, the streaming media file may essentially be an audio-video file that is distributed through streaming. The audio-video stream may include consecutive frames. The consecutive frames may include key frames and intermediate frames. Each frame may include a streaming timestamp and a duration to indicate a starting time to play in the audio-video stream and a playing length.

Slicing the streaming media file by the file processing module 21 into slices according to a preset time unit may include obtaining a duration of each frame in the streaming media file, and obtaining every slice by slicing the streaming media file sequentially based on the frame durations and the preset time unit.

The preset time unit may be set according to actual requirements. The preset time unit may be a specific value or a range of values, not limited by the present disclosure. It should be understood that, because the duration of each frame is determined according to the specific audio-video stream, in the process of slicing, the boundary may not correspond to a frame perfectly. In this case, a rounding method may be set forth to take care of such scenario. For example, a frame closest to the boundary may be selected as the ending frame of the slice. No limitation is imposed by the present disclosure

At the same time as the file processing module 21 slices and obtains the streaming media file, the file processing module 21 may create the corresponding playing sequence numbers based on time data.

Specifically, creating a playing sequence number by the file processing module 21 for each slice may include applying time data into a function to calculate a playing sequence number. The time data may include the streaming timestamp for the first frame in the current slice. For example, in one embodiment, the calculation function may be: Fx=tx−tx−1+Fx−1, where x is the sequence number of a slice, F(x) is the playing sequence number for the slice x, t(x) is the streaming timestamp of the first frame in the slice x, and F(0) is the current system time.

The created playing sequence number may be used to determine an order of playing the slice. The slice playing sequence number created by the file processing module 21 may intuitively reflect the playing order and specific streaming timestamp of the slices. In the subsequent playing process, the playing sequence numbers may be directly used to play the slices. Further, the playing sequence numbers may be used to detect abnormality, which will be described in detail later.

In certain aspects, the calculation function may be perfected to include weights and a system tolerance offset such that the created playing sequence numbers may be more realistic. For example, the calculation function may be Fx=Atx−Btx−1+Fx−1+C, where A and B are weights, and C is a system tolerance offset. The specific values of A, B, and C may be adjusted according to actual scenario of the system design. No limitation is imposed by the present disclosure.

It should be noted that, the present disclosure may use other methods to create the playing sequence numbers. For example, the streaming timestamps may be directly used to create the playing sequence numbers. Other time information may be used as long as the created playing sequence numbers reflect the real-time relationship between each of the slices.

Because the transmission of the streaming media file may be affected by network fluctuations and loss of data may occur, when the occurrence of loss of data is not detected timely, the playing sequence number creation may be affected such that, in the subsequent playing process, the playing sequence numbers may not be obtained correctly.

In one aspect, when creating a playing sequence number, the file processing module 21 may first determine whether a calibration is required. When it is determined that a calibration is required, the playing sequence number of the current slice may be created according to the current system time. When it is determined that a calibration is not required, the playing sequence number of the current slice may be created as previously described.

Determining whether a calibration is required includes determining whether loss of data occurs between the current slice and the last slice. When loss of data occurs, a calibration may be required.

Specifically, the streaming timestamp of frame may be used to determine whether loss of data occurs. Under normal circumstances, the streaming timestamp difference between adjacent frames of adjacent slices may be the duration of the preceding frame. Frame durations may often be within a reasonable range. By determining whether the streaming timestamp difference between the first frame of the current slice and the last frame of the preceding slice is within a reasonable range, whether loss of data occurs between the current slice and the preceding slice may be determined. In other embodiments, whether loss of data occurs may be determined by determining whether the streaming timestamp difference between the first frames of adjacent slices is within a reasonable range.

After it is determined that the loss of data has occurred, when the playing sequence number is still calculated based on the original method, a gap may appear between the playing sequence number of the current slice and the playing sequence number of the preceding slice. In the subsequent playing process, a request may fall into the gap, and no resource may be available to respond to the request. Thus, after it is determined that the loss of data has occurred, a calibration may be required for creating the playing sequence number to fill the gap and to avoid the scenario that no resource is available to respond to certain request. Because the creation of the playing sequence number is based on the system time and the slice playing time, the current system time may be directly used to calibrate the current playing sequence number.

The storage module 22 may receive data sent from the file processing module 21, store the slices in a first section 22a, and store the corresponding playing sequence numbers and mapping relationship in a second section 22b.

The player 30 may be used to play the streaming media file. In a specific live streaming scenario, the player 30 is a client terminal. After a user requests for a live streaming program, the player 30 may obtain a playing resource from the server 20 according to the user request.

Referring to FIG.6 and FIG.5 simultaneously, FIG. 6 illustrates a schematic diagram of a player of an exemplary live streaming system according to the present disclosure. As shown in FIG. 5 and FIG. 6, the player 30 may include a request unit 31 and a playing unit 32.

The request unit 31 may retrieve and select the playing sequence number of the user requested program from the second section 22b of the storage module 22 on the server 20 based on the user request, download the slice for playing from the first section 22a of the storage module 22 based on the playing sequence number and the mapping relationship, and send the downloaded slice to the playing unit 32 for playing.

In one aspect, the request unit 31 may retrieve the latest playing sequence number from the second section 22b. The latest playing sequence number may be the most recently created and stored playing sequence number. The player 30 may determine whether pausing occurs based on the playing sequence number of the currently played slice and the latest playing sequence number. When it is determined that the pausing has occurred, the player 30 may determine a next playing sequence number based on the playing delay and the latest playing sequence number. When it is determined that no pausing has occurred, the player 30 may determine a next playing sequence number based on the order of the playing sequence number of the second section 22b, that is, the playing sequence number immediately following the playing sequence number of the currently played slice.

Specifically, based on the transmission characteristics of the streaming media file, the player 30 may download next slice for playing at the same time as the current slice is played. In an ideal network environment, the most recently created playing sequence number may be the playing sequence number of the slice for playing next. However, in a real life network environment, the transmission network may have fluctuations such that the player may not be able to obtain resource timely, and image pausing may occur.

To address the issue, the player 30 may first determine whether the current streaming is pausing to decide the playing sequence number of the slice for playing next. Specifically, determining whether pausing occurs based on the playing sequence number of the currently played slice and the latest playing sequence number may include calculating a numerical difference between the playing sequence number of the currently played slice and the latest playing sequence number, and determining whether pausing occurs based on whether the numerical difference is within a preset range.

Because the playing sequence numbers are created based on time data, the playing sequence numbers may be directly used to determine a playing time for a slice. Theoretically, the numerical difference between the playing sequence number of the currently played slice and the latest playing sequence number may be equal to the playing time of the currently played slice, that is, the preset time unit when the streaming media file is sliced. Further, inevitable playing delay and system tolerance may be accommodated. When the numerical difference is within a reasonable preset range (e.g., the playing time of the slice + playing delay + system tolerance), the currently played slice may be played successfully, and no pausing may occur. Otherwise, the occurrence of pausing may be determined. The playing delay may be the delay in time between the image played at the client and the real-time image at the streaming host due to the time required for processing and transmitting data from the streaming host to the client in the live streaming in addition to the delay caused by the equipment reason of the player itself.

When it is determined that no pausing occurred, the player 30 may directly determine that the next playing sequence number is the sequence number immediately following the currently played sequence number in the record.

When it is determined that the pausing has occurred, in order to ensure the real-time playing of the images received by the user, the player 30 may skip the delayed slices due to pausing to ensure the real-time playing of the current images. Thus, the slice for playing next may be adjusted. The next playing sequence number may be determined based on the playing delay and the latest playing sequence number. Further, a system tolerance may be accommodated, that is, the next playing sequence number is equal to the latest playing sequence number minus a sum of the playing delay and the system tolerance.

In one aspect, after the file processing module 21 slices the streaming media file into slices, each slice that includes a key frame may be recorded. When pausing occurs and image skipping is required, after the next playing sequence number is determined, the player 30 may determine again whether the slice corresponding to the playing sequence number includes a key frame. When it is determined that no key frame is included, a playing sequence number of an adjacent slice that includes a key frame may be selected to be the next playing sequence number. Thus, snowy screen symptom caused by jumping to a slice that does not include a key frame may be avoided.

In the live streaming system as shown in FIG. 4, the data transmission between the push-stream terminal, the server, and the player and the respective functions may be illustrated. In a live streaming, the push-stream terminal may continuously generate the audio-video file, and may transfer the audio-video streaming media file to the server in real-time. The server may receive the streaming media file from the push-stream terminal, slice the streaming media file into slices, and create a playing sequence number for each slice based on time information. When creating a playing sequence number, the server may first determine whether loss of data occurs. When it is determined that loss of data has occurred, the playing sequence number may be adjusted by the current system time to avoid the situation that the player requests the lost data and gets no response. The player may sequentially request and download the slices from the server, and may play the downloaded slices. Each time when requesting and downloading a slice, the player may determine whether pausing occurs based on the playing sequence number. When it is determined that the pausing has occurred, the player may timely skip to avoid extensive playing delay that makes the playing not in real-time.

FIG. 7 illustrates a schematic diagram of another exemplary live streaming system according to the present disclosure. As shown in FIG. 7, the live streaming system 2 may include a push-stream terminal 100, a plurality of servers 200 (200a, 200b, ..., 200n as shown in the drawing), and a plurality of players 300 (300al, 300a2, ..., 300bl, 300b2, ..., as shown in the drawing).

The server 200a, 200b, ..., 200n may be connected to the push-stream terminal 100. The players 300al, 300a2 may be connected to the server 200a. The player 300b1, 300b2 may be connected to the server 200b. In one embodiment, the players may obtain playing resources through the connections to the corresponding servers. The servers may obtain the program resources of the push-stream terminal through the connections to the push-stream terminal. The push-stream terminal, the servers, and the players may have identical structures and functions as the push-stream terminal, the server, and the player as shown in FIG. 4. Thus, the description is not repeated.

In other aspects, the relationship between the push-stream terminals and the servers and the relationship between the servers and the players may be a many-to-many relationship. That is, one push-stream terminal maybe connected to multiple servers, one server may be connected to multiple push-stream terminals, or one player may be connected to multiple servers. The specific arrangements may be determined according to the actual business requirement and equipment configuration.

Through the descriptions of various aforementioned aspects, those skilled in the art may clearly understand that the aspects may be implemented by means of software in conjunction with an essential common hardware platform, or may be simply implemented by hardware. Based on such understanding, the essential part of the aforementioned technical solutions or the part that contribute to the prior art may be embodied in the form of software products. The software products may be stored in computer readable storage media, such as ROM/RAM, magnetic disk, and optical disk, etc., and may include a plurality of instructions to enable a computer device (may be a personal computer, a server, or a network device) having one or more processors to execute the methods described in various embodiments or parts of the embodiments.

The scope of the invention is defined by the appended claims.

 A streaming media file processing method, comprising: slicing (S10) a streaming media file into slices according to a preset time unit; creating (S20) a playing sequence number for each slice based on time data, wherein the playing sequence number is configured to determine an order of playing the slices; and storing (S30) the slices and the playing sequence numbers of the slices separately and establishing a mapping relationship between the slices and the corresponding playing sequence numbers for playing the steaming media file. The method according to claim 1, wherein slicing the streaming media file into slices according to the preset time unit includes: obtaining a duration for each frame in the streaming media file; and based on the duration and the preset time unit, obtaining the slice by slicing the streaming media file. The method according to claim 1, wherein:
the time data includes a streaming timestamp of a first frame in the slice, and creating a playing sequence number for the slice based on the time data includes:
applying the time data into a function to calculate the playing sequence number based on the time data. The method according to claim 1, wherein:
the time data includes a streaming timestamp of a first frame in the slice, and creating a playing sequence number for the slice further includes: determining whether a calibration is required for a currently created playing sequence number; when it is determined that a calibration is required, creating the playing sequence number for the current slice according to the system time; and when it is determined that no calibration is required, applying the time data into a predetermined function to calculate the playing sequence number. The method according to claim 4, wherein determining whether calibration is required for currently created playing sequence number includes: determining whether loss of data occurs between a current slice and a preceding slice; and when it is determined that loss of data has occurred, determining that a calibration is required. The method according to claim 3 or 4, wherein playing the steaming media file includes: retrieving (S101) a latest playing sequence number; determining (S102) whether pausing occurs based on a playing sequence number of the currently played slice and the latest playing sequence number; when it is determined that the pausing has occurred, determining a next playing sequence number based on a playing delay and the latest playing sequence number; when it is determined that no pausing has occurred, determining the next playing sequence number based on the order of playing sequence numbers; and downloading (S103) a slice corresponding to the next playing sequence number. The method according to claim 6, wherein determining whether pausing occurs includes: calculating a numerical difference between the playing sequence number of the currently played slice and the latest playing sequence number; and determining whether pausing occurs by determining whether the numerical difference is within a preset range. A live streaming system, comprising a push-stream terminal (10), a server (20), and a player (30), wherein: the push-stream terminal transfers a streaming media file to the server; a file processing module (21) of the server receives the streaming media file, slices the received streaming media file into slices according to a present time unit, creates a playing sequence number for each slice based on time data, establishes mapping relationship between the slices and the corresponding playing sequence numbers for playing the steaming media file, and respectively sends the slices, the playing sequence numbers of the slices, and the mapping relationship to a storage module (22) of the server (20); the storage module stores the slices in a first section (22a), and stores separately from the slices the playing sequence numbers of the slices and the mapping relationship in a second section (22b); and the player obtains and selects the playing sequence number from the second section, and downloads and plays a slice from the first section based on the playing sequence number and the mapping relationship. The live streaming system according to claim 8, wherein slicing the streaming media file into slices by the file processing module of the server according to a present time unit includes: obtaining a duration for each frame in the steaming media file; and obtaining the slice by slicing the streaming media file based on the frame durations and the preset time unit. The live streaming system according to claim 8, wherein:
the time data includes a streaming timestamp of a first frame in the slice, and creating a playing sequence number for the slice by the file processing module of the server based on time data includes:
applying the time data into a predetermined function to calculate the playing sequence number based on the time data. The live streaming system according to claim 10, wherein:
the time data includes a streaming timestamp of a first frame in the slice, and creating a playing sequence number for the slice by the file processing module of the server based on time data further includes: determining whether a calibration is required for a currently created playing sequence number; when it is determined that a calibration is required, creating the playing sequence number for the current slice according to the system time; and when it is determined that no calibration is required, applying the time data into a predetermined function to calculate the playing sequence number. The live streaming system according to claim 11, wherein determining by the file processing module whether a calibration is required for the currently created playing sequence number includes: determining whether loss of data occurs between a current slice and a preceding slice; and when it is determined that loss of data has occurred, determining that a calibration is required. The live streaming system according to claim 10 or 11, wherein: the player includes a request unit (31) and a playing unit (32); after a next playing sequence number is determined, the request unit sends a request to the storage module of the server to download the next slice and sends the downloaded slice to the playing unit to play; and when the request unit determines the playing sequence number for the requested slice, the request unit first obtains the latest playing sequence number, and then determines whether pausing occurs based on the playing sequence number of the currently played slice and the latest playing sequence number; when it is determined that the pausing has occurred, the request unit determines a next playing sequence number based on a playing delay and the latest playing sequence number; when it is determined that no pausing has occurred, the request unit determines a next playing sequence number based on the order of the playing sequence numbers, and sends a request to the server to download the slice corresponding to the next playing sequence number. The live streaming system according to claim 13, wherein:
the request unit calculates a numerical difference between the playing sequence number of the currently played slice and the latest playing sequence umber, and determines whether pausing occurs based on whether the numerical difference is within a preset range. The live streaming system according to claim 8, wherein:
after the file processing module slices the streaming media file into slices, whether each of the slices includes a key frame is recorded. Streaming-Media-Dateiverarbeitungsverfahren, umfassend: Slicen (S10) einer Streaming-Media-Datei in Slices gemäß einer voreingestellten Zeiteinheit; Erzeugen (S20) einer Abspielsequenznummer für jede Slice auf der Basis von Zeitdaten, wobei die Abspielsequenznummer dazu konfiguriert ist, eine Reihenfolge des Abspielens der Slices zu bestimmen; und separates Speichern (S30) der Slices und der Abspielsequenznummern der Slices und Erstellen einer Abbildungsbeziehung zwischen den Slices und den entsprechenden Abspielsequenznummern zum Abspielen der Streaming-Media-Datei. Verfahren nach Anspruch 1, wobei das Slicen der Streaming-Media-Datei in Slices gemäß der voreingestellten Zeiteinheit umfasst: Erhalten einer Dauer für jeden Frame in der Streaming-Media-Datei und auf der Basis der Dauer und der voreingestellten Zeiteinheit Erhalten der Slice durch Slicen der Streaming-Media-Datei. Verfahren nach Anspruch 1, wobei:
die Zeitdaten einen Streaming-Zeitstempel eines ersten Frames in der Slice umfassen und das Erzeugen einer Abspielsequenznummer für die Slice auf der Basis der Zeitdaten umfasst:
Einbringen der Zeitdaten in eine Funktion, um die Abspielsequenznummer auf der Basis der Zeitdaten zu berechnen. Verfahren nach Anspruch 1, wobei:
die Zeitdaten einen Streaming-Zeitstempel eines ersten Frames in der Slice umfassen und das Erzeugen einer Abspielsequenznummer für die Slice weiterhin umfasst: Bestimmen, ob eine Kalibrierung für eine aktuell erzeugte Abspielsequenznummer erforderlich ist; wenn bestimmt wird, dass eine Kalibrierung erforderlich ist, Erzeugen der Abspielsequenznummer für die aktuelle Slice gemäß der Systemzeit; und wenn bestimmt wird, dass keine Kalibrierung erforderlich ist, Einbringen der Zeitdaten in eine vorherbestimmte Funktion, um die Abspielsequenznummer zu berechnen. Verfahren nach Anspruch 4, wobei das Bestimmen, ob eine Kalibrierung für eine aktuell erzeugte Abspielsequenznummer erforderlich ist, umfasst: Bestimmen, ob ein Datenverlust zwischen einer aktuellen Slice und einer vorangehenden Slice auftritt; und wenn bestimmt wird, dass ein Datenverlust aufgetreten ist, Bestimmen, dass eine Kalibrierung erforderlich ist. Verfahren nach Anspruch 3 oder 4, wobei das Abspielen der Streaming-Media-Datei umfasst: Abrufen (S101) einer neuesten Abspielsequenznummer; Bestimmen (S102), ob ein Pausieren auftritt, auf der Basis einer Abspielsequenznummer der aktuell abgespielten Slice und der neuesten Abspielsequenznummer; wenn bestimmt wird, dass ein Pausieren aufgetreten ist, Bestimmen einer nächsten Abspielsequenznummer auf der Basis einer Abspielverzögerung und der neuesten Abspielsequenznummer; wenn bestimmt wird, dass kein Pausieren aufgetreten ist, Bestimmen der nächsten Abspielsequenznummer auf der Basis der Reihenfolge von Abspielsequenznummern und Herunterladen (S103) einer Slice, die der nächsten Abspielsequenznummer entspricht. Verfahren nach Anspruch 6, wobei das Bestimmen, ob ein Pausieren auftritt, umfasst: Berechnen einer numerischen Differenz zwischen der Abspielsequenznummer der aktuell abgespielten Slice und der neuesten Abspielsequenznummer und Bestimmen, ob ein Pausieren auftritt, durch Bestimmen, ob die numerische Differenz innerhalb eines voreingestellten Bereichs liegt. Live-Streaming-System, umfassend ein Push-Stream-Endgerät (10), einen Server (20) und einen Player (30), wobei: das Push-Stream-Endgerät eine Streaming-Media-Datei an den Server überträgt; ein Dateiverarbeitungsmodul (21) des Servers die Streaming-Media-Datei empfängt, die empfangene Streaming-Media-Datei in Slices gemäß einer voreingestellten Zeiteinheit slicet, eine Abspielsequenznummer für jede Slice auf der Basis von Zeitdaten erzeugt, eine Abbildungsbeziehung zwischen den Slices und den entsprechenden Abspielsequenznummern zum Abspielen der Streaming-Media-Datei erstellt und die Slices, die Abspielsequenznummern der Slices bzw. die Abbildungsbeziehung an ein Speichermodul (22) des Servers (20) sendet; das Speichermodul die Slices in einem ersten Abschnitt (22a) speichert und die Abspielsequenznummern der Slices und die Abbildungsbeziehung separat von den Slices in einem zweiten Abschnitt (22b) speichert und der Player die Abspielsequenznummer von dem zweiten Abschnitt erhält und auswählt und eine Slice von dem ersten Abschnitt auf der Basis der Abspielsequenznummer und der Abbildungsbeziehung herunterlädt und abspielt. Live-Streaming-System nach Anspruch 8, wobei das Slicen der Streaming-Media-Datei in Slices durch das Dateiverarbeitungsmodul des Servers gemäß einer voreingestellten Zeiteinheit umfasst: Erhalten einer Dauer für jeden Frame in der Streaming-Media-Datei und Erhalten der Slice durch Slicen der Streaming-Media-Datei auf der Basis der Frame-Dauern und der voreingestellten Zeiteinheit. Live-Streaming-System nach Anspruch 8, wobei:
die Zeitdaten einen Streaming-Zeitstempel eines ersten Frames in der Slice umfassen und das Erzeugen einer Abspielsequenznummer für die Slice durch das Dateiverarbeitungsmodul des Servers auf der Basis von Zeitdaten umfasst:
Einbringen der Zeitdaten in eine vorherbestimmte Funktion, um die Abspielsequenznummer auf der Basis der Zeitdaten zu berechnen. Live-Streaming-System nach Anspruch 10, wobei:
die Zeitdaten einen Streaming-Zeitstempel eines ersten Frames in der Slice umfassen und das Erzeugen einer Abspielsequenznummer für die Slice durch das Dateiverarbeitungsmodul des Servers auf der Basis von Zeitdaten weiterhin umfasst: Bestimmen, ob eine Kalibrierung für eine aktuell erzeugte Abspielsequenznummer erforderlich ist; wenn bestimmt wird, dass eine Kalibrierung erforderlich ist, Erzeugen der Abspielsequenznummer für die aktuelle Slice gemäß der Systemzeit; und wenn bestimmt wird, dass keine Kalibrierung erforderlich ist, Einbringen der Zeitdaten in eine vorherbestimmte Funktion, um die Abspielsequenznummer zu berechnen. Live-Streaming-System nach Anspruch 11, wobei das Bestimmen durch das Dateiverarbeitungsmodul, ob eine Kalibrierung für die aktuell erzeugte Abspielsequenznummer erforderlich ist, umfasst: Bestimmen, ob ein Datenverlust zwischen einer aktuellen Slice und einer vorangehenden Slice auftritt; und wenn bestimmt wird, dass ein Datenverlust aufgetreten ist, Bestimmen, dass eine Kalibrierung erforderlich ist. Live-Streaming-System nach Anspruch 10 oder 11, wobei: der Player eine Anforderungseinheit (31) und eine Abspieleinheit (32) umfasst; nachdem eine nächste Abspielsequenznummer bestimmt wurde, die Anforderungseinheit eine Anforderung an das Speichermodul des Servers sendet, um die nächste Slice herunterzuladen, und die heruntergeladene Slice an die Abspieleinheit zum Abspielen sendet; und wenn die Anforderungseinheit die Abspielsequenznummer für die angeforderte Slice bestimmt, die Anforderungseinheit zunächst die neueste Abspielsequenznummer erhält und dann bestimmt, ob ein Pausieren auftritt, auf der Basis der Abspielsequenznummer der aktuell abgespielten Slice und der neuesten Abspielsequenznummer; wenn bestimmt wird, dass das Pausieren aufgetreten ist, die Anforderungseinheit eine nächste Abspielsequenznummer auf der Basis einer Abspielverzögerung und der neuesten Abspielsequenznummer bestimmt; wenn bestimmt wird, dass kein Pausieren aufgetreten ist, die Anforderungseinheit eine nächste Abspielsequenznummer auf der Basis der Reihenfolge der Abspielsequenznummern bestimmt und eine Anforderung an den Server sendet, um die Slice herunterzuladen, die der nächsten Abspielsequenznummer entspricht. Live-Streaming-System nach Anspruch 13, wobei:
die Anforderungseinheit eine numerische Differenz zwischen der Abspielsequenznummer der aktuell abgespielten Slice und der neuesten Abspielsequenznummer berechnet und bestimmt, ob ein Pausieren auftritt, auf der Basis davon, ob die numerische Differenz innerhalb eines voreingestellten Bereichs liegt. Live-Streaming-System nach Anspruch 8, wobei:
nachdem das Dateiverarbeitungsmodul die Streaming-Media-Datei in Slices gesclicet hat, aufgezeichnet wird, ob jede der Slices einen Schlüssel-Frame umfasst. Procédé de traitement de fichier multimédia de diffusion en continu, comprenant : découper (S10) un fichier multimédia de diffusion en continu en tranches selon une unité temporelle préréglée ; créer (S20) un numéro de séquence de lecture pour chaque tranche sur la base de données temporelles, le numéro de séquence de lecture étant configuré pour déterminer un ordre de lecture des tranches ; et stocker (S30) les tranches et les numéro de séquence de lecture des tranches séparément et établir une relation de mappage entre les tranches et les numéros de séquence de lecture correspondants pour la lecture du fichier multimédia de diffusion en continu. Procédé selon la revendication 1, dans lequel la découpe du fichier multimédia de diffusion en continu en tranches selon l'unité temporelle préréglée comprend : obtenir une durée pour chaque trame dans le fichier multimédia de diffusion en continu ; et sur la base de la durée et de l'unité temporelle préréglée, obtenir la tranche par découpe du fichier multimédia de diffusion en continu. Procédé selon la revendication 1, dans lequel :
les données temporelles comprennent une estampille temporelle de diffusion en continu d'une première trame dans la tranche, et la création d'un numéro de séquence de lecture pour la tranche sur la base des données temporelles comprend :
appliquer les données temporelles dans une fonction pour calculer le numéro de séquence de lecture sur la base des données temporelles. Procédé selon la revendication 1, dans lequel :
les données temporelles comprennent une estampille temporelle de diffusion en continu d'une première trame dans la tranche, et la création d'un numéro de séquence de lecture pour la tranche comprend en outre : déterminer si un étalonnage est ou non requis pour un numéro de séquence de lecture créé actuellement ; lorsqu'il est déterminé qu'un étalonnage est requis, créer le numéro de séquence de lecture pour la tranche actuelle selon le temps de système ; et lorsqu'il est déterminé qu'aucun étalonnage n'est requis, appliquer les données temporelles dans une fonction prédéterminée pour calculer le numéro de séquence de lecture. Procédé selon la revendication 4, dans lequel la détermination du point de savoir si un étalonnage est ou non requis pour un numéro de séquence de lecture créé actuellement comprend : déterminer si une perte de données se produit ou non entre une tranche actuelle et une tranche précédente ; et lorsqu'il est déterminé qu'une perte de données s'est produite, déterminer qu'un étalonnage est requis. Procédé selon la revendication 3 ou 4, dans lequel la lecture du fichier multimédia de diffusion en continu comprend : extraire (S101) un dernier numéro de séquence de lecture ; déterminer (S102) si une pause se produit ou non sur la base d'un numéro de séquence de lecture de la tranche lue actuellement et du dernier numéro de séquence de lecture ; lorsqu'il est déterminé que la pause s'est produite, déterminer un numéro de séquence de lecture suivant sur la base d'un retard de lecture et du dernier numéro de séquence de lecture ; lorsqu'il est déterminé qu'aucune pause ne s'est produite, déterminer le numéro de séquence de lecture suivant sur la base de l'ordre de numéros de séquence de lecture ; et télécharger (S103) une tranche correspondant au numéro de séquence de lecture suivant. Procédé selon la revendication 6, dans lequel la détermination du point de savoir si une pause se produit ou non comprend : calculer une différence numérique entre le numéro de séquence de lecture de la tranche lue actuellement et le dernier numéro de séquence de lecture ; et déterminer si une pause se produit ou non par détermination du point de savoir si la différence numérique est ou non dans une plage préréglée. Système de diffusion en continu en direct, comprenant un terminal de flux de diffusion sélective (10), un serveur (20) et un lecteur (30), dans lequel : le terminal de flux de diffusion sélective transfère un fichier multimédia de diffusion en continu vers le serveur ; un module de traitement de fichier (21) du serveur reçoit le fichier multimédia de diffusion en continu, découpe le fichier multimédia de diffusion en continu reçu en tranches selon une unité temporelle préréglée, crée un numéro de séquence de lecture pour chaque tranche sur la base de données temporelles, établit une relation de mappage entre les tranches et les numéros de séquence de lecture correspondants pour la lecture du fichier multimédia de diffusion en continu, et envoie respectivement les tranches, les numéros de séquence de lecture des tranches et la relation de mappage à un module de stockage (22) du serveur (20) ; le module de stockage stocke les tranches dans une première section (22a), et stocke séparément des tranches les numéros de séquence de lecture des tranches et la relation de mappage dans une seconde section (22b) ; et le lecteur obtient et sélectionne le numéro de séquence de lecture à partir de la seconde section, et télécharge et lit une tranche à partir de la première section sur la base du numéro de séquence de lecture et de la relation de mappage. Système de diffusion en continu en direct selon la revendication 8, dans lequel la découpe du fichier multimédia de diffusion en continu en tranches par le module de traitement de fichier du serveur selon une unité temporelle préréglée comprend : obtenir une durée pour chaque trame dans le fichier multimédia de diffusion en continu ; et obtenir la tranche par découpe du fichier multimédia de diffusion en continu sur la base des durées de trame et de l'unité temporelle préréglée. Système de diffusion en continu en direct selon la revendication 8, dans lequel :
les données temporelles comprennent une estampille temporelle de diffusion en continu d'une première trame dans la tranche, et la création d'un numéro de séquence de lecture pour la tranche par le module de traitement de fichier du serveur sur la base de données temporelles comprend :
appliquer les données temporelles dans une fonction prédéterminée pour calculer le numéro de séquence de lecture sur la base des données temporelles. Système de diffusion en continu en direct selon la revendication 10, dans lequel :
les données temporelles comprennent une estampille temporelle de diffusion en continu d'une première trame dans la tranche, et la création d'un numéro de séquence de lecture pour la tranche par le module de traitement de fichier du serveur sur la base de données temporelles comprend en outre : déterminer si un étalonnage est ou non requis pour un numéro de séquence de lecture créé actuellement ; lorsqu'il est déterminé qu'un étalonnage est requis, créer le numéro de séquence de lecture pour la tranche actuelle selon le temps de système ; et lorsqu'il est déterminé qu'aucun étalonnage n'est requis, appliquer les données temporelles dans une fonction prédéterminée pour calculer le numéro de séquence de lecture. Système de diffusion en continu en direct selon la revendication 11, dans lequel la détermination par le module de traitement de fichier du point de savoir si un étalonnage est ou non requis pour le numéro de séquence de lecture créé actuellement comprend : déterminer si une perte de données se produit ou non entre une tranche actuelle et une tranche précédente ; et lorsqu'il est déterminé qu'une perte de données s'est produite, déterminer qu'un étalonnage est requis. Système de diffusion en continu en direct selon la revendication 10 ou 11, dans lequel : le lecteur comprend une unité de requête (31) et une unité de lecture (32) ; après qu'un numéro de séquence de lecture suivant est déterminé, l'unité de requête envoie une requête au module de stockage du serveur pour télécharger la tranche suivante et envoie la tranche téléchargée à l'unité de lecture en vue d'une lecture ; et lorsque l'unité de requête détermine le numéro de séquence de lecture pour la tranche requise, l'unité de requête tout d'abord obtient le dernier numéro de séquence de lecture, puis détermine si une pause se produit ou non sur la base du numéro de séquence de lecture de la tranche lue actuellement et du dernier numéro de séquence de lecture ; lorsqu'il est déterminé que la pause s'est produite, l'unité de requête détermine un numéro de séquence de lecture suivant sur la base d'un retard de lecture et du dernier numéro de séquence de lecture ; lorsqu'il est déterminé qu'aucune pause ne s'est produite, l'unité de requête détermine un numéro de séquence de lecture suivant sur la base de l'ordre des numéros de séquence de lecture, et envoie une requête au serveur pour télécharger la tranche correspondant au numéro de séquence de lecture suivant. Système de diffusion en continu en direct selon la revendication 13, dans lequel :
l'unité de requête calcule une différence numérique entre le numéro de séquence de lecture de la tranche lue actuellement et le dernier numéro de séquence de lecture, et détermine si une pause se produit ou non sur la base du point de savoir si la différence numérique est ou non dans une plage préréglée. Système de diffusion en continu en direct selon la revendication 8, dans lequel :
après que le module de traitement de fichier découpe le fichier multimédia de diffusion en continu en tranches, le point de savoir si chacune des tranches comprend ou non une trame clé est enregistré.
REFERENCES CITED IN THE DESCRIPTION

This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Patent documents cited in the description